The naming of parts Ontologies and controlled vocabularies
The naming of parts: Ontologies and controlled vocabularies.
What are ontologies and controlled vocabularies? 'When I use a word, ' Humpty Dumpty said, in a rather scornful tone, ' it means just what I choose it to mean, neither more nor less. ’ 'The question is, ' said Alice, 'whether you can make words mean so many different things. ' 'The question is, ' said Humpty Dumpty, 'which is to be master that's all’. Alice through the looking glass - Lewis Carroll 1871. Ontologies are the right word in the right place
Choose your words carefully "What's in a name? That which we call a rose By any other word would smell as sweet. " -- Romeo and Juliet (II, ii, 1 -2) Plant ontologies at NASC Hairy Hirsute Pubescent Bald Smooth Glabrous
Ontologies are primarily used for disambiguity Homonym horrors: Their They’re There Metal tolerance
But must be practical / pragmatic / fit for purpose Rule 1. of ontologies – the users must be willing to use them Simple functional Classifications are most acceptable
The more complex / heterogeneous, the harder they are to use ………and then they simply aren’t…. . (used). Remember - bio-ontologies are defined groupings Not natural or philosophical absolutes.
Curation sometimes depends on description Strong visual phenotypes can often be reasonably represented by photo(s) Visually cryptic – e. g. molecular phenotypes are not
Molecular phenotype ontologies include Gene Ontology (GO) hierarchical
Molecular ontologies allow for machine techniques – e. g. Gene set enrichment (GO enrichment) Enrichment map in Cytoscape Much of Bioinformatics is analysing data, storing it sensibly - and making it easy(ier) to assess. E. g. reduction of dimensionality to aid visualisation: Network graphs, PCAs, GUIs. Because humans are good at patterns but bad at numbers (especially big ones).
Plant phenotype ontologies are concatenated controlled vocabularies • Entity, (structure) Taken from Plant Ontology ( PO ) - e. g leaf, microspore. …modified by properties of that entity (size, shape, colour, etc): • Attribute and Value (EAV) -or • Quality (EQ) Taken from the Phenotype, Attribute and Trait Ontology ( PATO ). e. g. Phenotype description in free text: Green dwarf. Broader leaves, glabra. Yellow seed.
Translated into catalogue curation
Defined structures allow constrained searches
…to discover stocks in a phenotype driven sub-category
Phenotype and ontologies practical Curation is often a steep learning curve for both users and IT staff. . Exercise 1: naïve curation
Task one – immediate description: Describe these four plants, spend ~ two minutes on each. N 89 N 3117 N 258 N 933
2. Brainstorm naïve curational Issues. Familiarity, context, names, opinion…. Some terms travel crossspecies: proximal/distal, medial/lateral. Some need translation: e. g. Rostral (nose) caudal (tail) in humans.
3. Enter plant structure (entity) and description (attribute value), separating with commas. Any entity/attribute values not present in the key can be added in the free text column
Results - improvement? Specialisation blindness ? (https: //youtu. be/v. JG 698 U 2 Mvo) What about automatic curation? Machine recognition, drones, IR, phenomics… Potential issue: Even within one species, structures can be quite diverse - so homologous structures between species may be very divergent. Exercise 4. Odd ones out How many species?
‘Standard’ Arabidopsis Mutants ‘test’standards Turnip Swede Pak Choi Single gene mutants Romanescu Double mutant: Ap 1/Cal . . and don’t forget varieties / landraces / ecotypes……. . etc. .
- Slides: 19